Comments on XACML 1.1 draft 2

[Satoshi Hada <SATOSHIH@jp.ibm.com>]

Hi, 

I'd like to restate my comments according
to the new draft posted at
http://lists.oasis-open.org/archives/xacml/200307/msg00033.html

(1) 
Section 5.4 (Lines 2032--2033): 
"The <XPathVersion> element SHALL specify 
the version of the XPath specification to be 
used by <AttributeSelector> elements." 

should be 

"The <XPathVersion> element SHALL specify 
the version of the XPath specification to be 
used by <AttributeSelector> elements and XPath-based functions."

This comment is based on Section 3.9 of the errata document.

(2) 
Two different terms are used to refer to a URI equality comparison.
"URI equality" in Sections 5.27 and 7.9.1, but "anyURI-equality"
in Section 5.28. 
We should have a common term. 

(3) 
According to the XACML schema, the data type of the "Issuer"
attribute 
in AttributeDesignatorType is xs:string. 
However,7.9.1 says that:
If Issuer is supplied in the named attribute,
then it MUST match, by URI equality, 
the Issuer of the same context attribute.
 

"URI equality" should be "string
equality". 

The comment about the "Issuer" attribute 
should go to the errata document because
the inconsistency is in the 1.0 specification. 

Satoshi Hada
IBM Tokyo Research Laboratory 
mailto:satoshih@jp.ibm.com
----- Forwarded by Satoshi
Hada/Japan/IBM on 2003/07/15 22:49 -----

Satoshi Hada/Japan/IBM@IBMJP 2003/07/15 15:44

To:        xacml@lists.oasis-open.org         cc:                 Subject:        [xacml] Comments on XACML 1.1 draft 1

Hi, 

I'm sending my comments on the draft of version 1.1 posted at 
http://lists.oasis-open.org/archives/xacml/200307/msg00001.html

(1) 
Section 5.4 (Lines 1840--1841): 
"The <XPathVersion> element SHALL specify 
the version of the XPath specification to be 
used by <AttributeSelector> elements." 

should be 

"The <XPathVersion> element SHALL specify 
the version of the XPath specification to be 
used by <AttributeSelector> elements and XPath-based functions."

This comment is based on Section 3.9 of the errata document.

(2) 
Two different terms are used to refer to a URI equality comparison.
"URI equality" in Sections 5.27 and 7.9.1, but "anyURI-equality"
in Section 5.28. 
We should have a common term. 

(3) 
According to the XACML schema, the data type of the "Issuer"
attribute 
in AttributeDesignatorType is xs:string. 
However, Sections 5.27 and 7.9.1 seem to say that it is xs:anyURI.
It should be consistent. 

This comment should go to the errata document because 
this is an inconsistency in the 1.0 specification.

Also, see 
http://lists.oasis-open.org/archives/xacml/200307/msg00011.html

Satoshi Hada
IBM Tokyo Research Laboratory 
mailto:satoshih@jp.ibm.com